Group level SEM relationships on soil moisture and external factors

[Kwabena A. Nketia]

Dear All,

I am working on soil moisture and trying to used SEM to understand the relationships and interaction of my data.

 

I used this model (below) but gets this error that I am unable to go around. Can you please assist. I have an extract of my large dataset for your use.

 

model <- '

                # Latent variables (measurement models)

                  Soil  =~ Clay + Depth + BD + Silt

                  Env   =~ TWI + LS

                  Clim  =~ API + ETo

 

                # Regressions

                  Soil ~ Env

                  Env ~ Clim

                  Grav ~ Soil + Env + Clim

                 

                # Residual correlation (Covariance)

                  Clay ~~ Depth

 

                # Intercept of observed varibales

                  Grav ~ 1

                  Clay ~ 1

                  Depth ~ 1

                  Silt ~ 1

                  BD ~ 1

   fit1 <- sem(model, data = SEM.df)

   fit2 <- sem(model, data = SEM.df, group = "SoilSeries")

 

 

error “  In lavaan::lavaan(model = model, data = SEM.df, model.type = "sem",  :

  lavaan WARNING: the optimizer warns that a solution has NOT been found! ”

Many thanks

Kwabena

[Terrence Jorgensen]

error “  In lavaan::lavaan(model = model, data = SEM.df, model.type = "sem",  :

  lavaan WARNING: the optimizer warns that a solution has NOT been found! ”

Are you using the latest lavaan (version 0.6-3)?  I do not get an error message, but I get the warning message above, as well as a second warning:

lavaan WARNING: some observed variances are (at least) a factor 1000 times larger than others; use varTable(fit) to investigate

Following the instructions in the message, I see huge discrepancies in the scales of your variables, which makes it difficult for the optimizer to search for a solution.  You can rescale your variables to express them in a different order of magnitude, for example:

## interpret effect of changing ETo by one-tenth of a unit
SEM.df$ETo_10 <- SEM.df$ETo * 10
## interpret effect of changing Silt by 100 units
SEM.df$Silt_100 <- SEM.df$Silt / 100

Terrence D. Jorgensen

Postdoctoral Researcher, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

http://www.uva.nl/profile/t.d.jorgensen

[Kwabena A. Nketia]

Many thanks Terrence

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/d/optout.

--

Nketia, Kwabena Abrefa (M.Phil)
PhD Candidate, Physical Geography - Georg August University, Germany
E-Mail: kwabena-ab...@geo.uni-goettingen.de

Ghana 
Research Scientist 
Soil Genesis, Survey & Classification Div.
CSIR-Soil Research Institute, PMB. Kwadaso - Kumasi - Ghana
Cell: +233 (0) 20 8278770; Email: kan...@gmail.com